Winner of the New Statesman SPERI Prize in Political Economy 2016

New! Lecture on 23rd May at Bush House, 44-46 Aldwych on my book 'The Lies We Were Told' with discussion from Rachel Shabi and Aeron Davis. Book here.

Monday, 27 April 2015

Mediamacro myth 7: the strong recovery

It was perhaps inevitable that when output did start growing again in 2013, everyone would breathe a huge sigh of relief, and extol what was happening. But that time has passed, and yet mediamacro - with few exceptions - has not told people the truth about the recovery. What they have not said is that the recovery does not really deserve that name. This picture tells you why.

The first point is to stop talking about GDP, and start talking about GDP per head. An economy that grows because it has more people in it is not obviously a good or bad thing, and from the government’s point of view - given its (missed) net migration target - it represents a failure. It is GDP per head that determines living standards, which is what matters.

Now if you were on a journey, and in one part of your journey you were delayed by 10 minutes because of a traffic jam, you would be relieved when that jam came to an end, but would you call moving again a recovery? Surely you would only talk about a recovery if you made up for some of that lost time. As the chart shows, we have failed as yet to make up for any of the ground lost not just in the 2009 recession, but also ground lost as a result of fiscal austerity in 2010 and 2011.

So we have not really seen a recovery. Maybe the pessimists are right, and we will never recover any of that lost output, but still you do not call it a recovery.

I can put it another way. Quarterly growth in GDP per head since the beginning of 2013 has averaged about 2% at an annual rate. That is below the average growth rate since 1955. A recovery from a deep recession would have growth rates well above the long term average.

So current growth is unexceptional, and nothing to write home about. The half-truth here is that growth elsewhere has also been poor, but largely because other countries have also implemented premature fiscal austerity. (In terms of GDP per head, both the US and Japan have done better than the UK since the recession.) But even GDP per head may be giving us an overoptimistic picture about average living standards. I’m going to break my one chart rule for this series to add this from the ONS. It plots GDP per head, and Net National Disposable Income (NNDI). The first measures production per head before depreciation, whereas NNDI measures income per head after depreciation.

NNDI has not increased at all under the coalition government, and the reason is that while overseas agents have been receiving some of the income from UK production, domestic residents’ income from overseas production has not been increasing by as much. This is a key reason why the UK current account deficit has been increasing.

In a nutshell, the prosperity of the average citizen in this country has hardly increased over the period of this coalition government - a result that is totally unprecedented since at least WWII. As recoveries from recessions go, this does not seem like a recovery worthy of the name. Yet we keep being told by mediamacro that the Coalition’s strong card is its economic record!

Previous posts in this series



  1. The chart doesn't recover to the pre ressesion path levels after 75 81 or 92 either.

    1. Output does return to the trend path in all 3 cases.

    2. The red trend line goes from 1955 to the peak 2008. Then you have the gap that you are refering to between the blue line and the red line after 2008 and to the end of the chart. If you do the same treatment for 75 81 and 92 there is also a permanent gap between the blue line and the corresponding pre recession trend lines for those recessions.

    3. But that is not how I drew the graph. I chose a number for the growth in underlying potential: 2.25%. What alternative number could you choose that would be plausible for pre-2008 and suggest a recovery after 2010?

    4. You reflect in the post that maybe the pesermists are right we will never recover that lost output, the chart shows that is the case for 75 81 92.

      Have you noticed that the term recovery is not used with a rigourouse definintion in Economics discourse. It could mean a return to previouse peak level, or it could mean return to previouse growth rate, or it could mean a return to where the economy would have been had there not been a recession, or it could mean a recovery from a diminishing trend to a growing trend.

    5. Am I being dense? How could you draw a trend line for any of those years that fits the years prior?

      You don't draw a trend line to the peak, you draw it to the trend. It just so happens that the 2008 peak was after a particularly long period of "on trend" growth, surely?

      What growth figure would give the line you think it should be?

  2. If you've ever had to have stuff published in local newspapers (the training ground very often for Fleet Street hacks) you may have found (as I did) that submitted stories are very often published word for word as you write them. Occasionally you may find they have more time and alter a sentence here or there, but fundamentally they remain unchanged. With greater and greater demands on higher productivity, the once local, now National, hacks who learned their shortcuts in local media are most likely repeating them on a larger scale.

    So, if a Party PR guy sends out a missive that puts in the phrase "xyz's strong card is its economic record" after years of being told the pain of recession is all the fault of the previous lot - aided and abetted by a certain 'note left on a chair' at the Treasury by one of the previous lot - mediamacro will simply reprint the phrase, unchanged, again and again.

    Perhaps you could begin with the Oxford Mail and go from there?

  3. Great blogging. My two cents with a link to an interesting paper by Bill Martin:

  4. The inclusion of the 'economic effects' of prostitution and drug dealing in the GDP figures is a reason why we seem to have 'grown'. Prostitutes and drug dealers don't declare their income or pay income tax, although some sex workers do have online 'businesses'. (An EU idea, apparently, so it's not just the UK that does this.)

    Then we have the situation where the tax authorities are actually paying out taxes in the form of Working/Family Tax Credits to prop up people's wages. Let's not forget the subsidising of landlords via Housing Benefit for those in work!

    1. AFAIK the drugs and prostitution figures are backdated to 1997, but I could be wrong, the best I could find was this article:

      I'd appreciate clarification if anyone knows.

  5. Very interesting view, and clearly the linear trend line is a remarkably good fit up to the current decade.

    But couldn't a case be made that just as it's much more meaningful to look at the trend for GDP per head rather than GDP, this analysis would be even more relevant if we look at real terms GDP. The figures are easily available here

    The same graph created on inflation adjusted figures still shows the economy is lagging the trend before the financial crisis, but provides more evidence that the recovery is now in progress . Given the historically low inflation this seems logical.

    This approach isn't as headline grabbing as your analysis, but I think there is a good case it may be fairer.

    1. IXHT is current price GDP per head. I'm looking at real GDP per head (what they call chained volume: IHXW) - I should have made that clear in the chart. Real GDP is more relevant than current price GDP.

    2. Many thanks for the clarification, I thought it was odd to have excluded any reference to real GDP. Thanks for explaining the best method of applying this.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.